This is a proposal to study the cortical mechanisms of pain perception in human using the technique of somatosensory evoked potentials (SEPs) to painful and non-painful stimuli. Recent MRI/PET studies have indicated that somatosensory cortical areas SI and SII, as well as anterior cingulate cortex, are activated by painful stimulation. These methods, however, have several disadvantages: they are invasive, expensive, not widely available, and have poor temporal resolution. In contrast, the SEP method is non-invasive, inexpensive and has the best temporal resolution (better than 1 msec) of any method for assessing brain activity. We have recently identified a long-latency component in the SEP, maximal at midline scalp sites, which appears to be specific for painful stimuli and which can be distinguished by latency and topography from other components related to stimulus intensity and cognitive processes. This pain SEP has the potential to become an objective, non-invasive measure of pain processing by the human central nervous system. In this project, we will carry out the following studies. First, we will better characterize the pain SEP by recording from multiple channels and by confirming that it is specifically related to the perception of stimuli perceived as painful and not simply related to stimulus intensity. We will confirm that is amplitude is correlated with subjects' subjective pain ratings. We will examine shorter latency components recorded over lateral scalp sites for evidence of pain- specific components originating in somatosensory cortex. Using recently developed methods for identifying the sources of scalp-recorded SEPs, we will attempt to determine the intracranial generators of components associated with pain processing. Our working hypothesis is that the earlier portions of the SEA(<100-150 msec latency) will contain pain- related components which arise from somatosensory cortical regions and thus will show topographical changes with stimulation at different sites on the body. In contrast, we hypothesize that the longer latency pain component which we have identified in preliminary studies will be related to later (>200 msec) activation of anterior cingulate cortex and will not vary in its dipole localization with stimulation of different body sites.